Photosynthetic Characteristics of Smaller and Larger Cell Size-Fractioned Phytoplankton Assemblies in the Daya Bay, Northern South China Sea

Cell size of phytoplankton is known to influence their physiologies and, consequently, marine primary production. To characterize the cell size-dependent photophysiology of phytoplankton, we comparably explored the photosynthetic characteristics of piconano- (<20 µm) and micro-phytoplankton cell assemblies (>20 µm) in the Daya Bay, northern South China Sea, using a 36-h in situ high-temporal-resolution experiment. During the experimental periods, the phytoplankton biomass (Chl a) in the surface water ranged from 0.92 to 5.13 μg L−1, which was lower than that in bottom layer (i.e., 1.83–6.84 μg L−1). Piconano-Chl a accounted for 72% (mean value) of the total Chl a, with no significant difference between the surface and bottom layers. The maximum photochemical quantum yield (FV/FM) of Photosystem II (PS II) and functional absorption cross-section of PS II photochemistry (σPS II) of both piconano- and micro-cells assemblies varied inversely with solar radiation, but this occurred to a lesser extent in the former than in the latter ones. The σPS II of piconano- and micro-cell assemblies showed a similar change pattern to the FV/FM in daytime, but not in nighttime. Moreover, the fluorescence light curve (FLC)-derived light utilization efficiency (α) displayed the same daily change pattern as the FV/FM, and the saturation irradiance (EK) and maximal rETR (rETRmax) mirrored the change in the solar radiation. The FV/FM and σPS II of the piconano-cells were higher than their micro-counterparts under high solar light; while the EK and rETRmax were lower, no matter in what light regimes. In addition, our results indicate that the FV/FM of the micro-cell assembly varied quicker in regard to Chl a change than that of the piconano-cell assembly, indicating the larger phytoplankton cells are more suitable to grow than the smaller ones in the Daya Bay through timely modulating the PS II activity.


Introduction
Marine phytoplankton, a polyphyletically diverse group of unicellular primary producers, can produce~50 Pg C per year by photosynthesis [1]. The cell size of these photosynthetic organisms can span from~0.6 to >1000 µm in the equivalent spherical diameter from the smallest cyanobacteria to the largest diatoms, with over nine orders of magnitude in biovolume [2,3]. The cell size range of phytoplankton often endows them with advantages in varied marine environments [4]. For instance, small phytoplankton cells that have a high surface-to-volume ratio usually outcompete their large counterparts for growth-limiting resources, e.g., carbon, nitrogen and phosphorus, and thereby dominate in

Study Area
A high temporal resolution experiment was carried out over a period of 36 h with a 2 h interval measurement on 25-26 March 2021 (early spring) in a coastal site (22°35′ N, 114°32′ E) of the Daya Bay, northern South China Sea ( Figure 1). This bay is semi-enclosed, with a sprawling coastline of ~52 km, covering an area of ~600 km 2 with a mean depth of less than 10 m [29]. The tide in this bay is irregularly semidiurnal with surface waterresident time of ~3.2 d [30]. Owing to the ebb and flow of the tide, the depth of the sampling site varied from 3.5 to 4.5 m during the experimental period.

Sampling Protocol
Every 2 h, the natural seawater with phytoplankton assemblages was collected from surface (~0.2 m) and bottom layers (~3.5 m) of the sampling site using a clean 5 L plexiglass water sampler to determine the environmental factors, and biological and physiological parameters described as follows.

Chlorophyll a (Chl a) Measurement
Every 2 h, 500 mL seawater from the surface and bottom layers were sequentially filtered through 20 µm pore-sized nylon-net filters (25 mm, Merck Millipore Ltd., Carrigtwohill, USA) and 0.7 µm pore-sized GF/F filters. The filters were then wrapped in aluminum foil, immediately frozen and stored at −20 °C for later analysis. After returning to the laboratory, these filters with cell assemblies were extracted overnight in a 5 mL

Study Area
A high temporal resolution experiment was carried out over a period of 36 h with a 2 h interval measurement on 25-26 March 2021 (early spring) in a coastal site (22 • 35 N, 114 • 32 E) of the Daya Bay, northern South China Sea ( Figure 1). This bay is semi-enclosed, with a sprawling coastline of~52 km, covering an area of~600 km 2 with a mean depth of less than 10 m [29]. The tide in this bay is irregularly semidiurnal with surface water-resident time of 3.2 d [30]. Owing to the ebb and flow of the tide, the depth of the sampling site varied from 3.5 to 4.5 m during the experimental period.

Sampling Protocol
Every 2 h, the natural seawater with phytoplankton assemblages was collected from surface (~0.2 m) and bottom layers (~3.5 m) of the sampling site using a clean 5 L plexiglass water sampler to determine the environmental factors, and biological and physiological parameters described as follows.

Environment Measurements
Every 2 h, the temperature, salinity and pH in surface and bottom layers were measured with a multi-parameter water quality monitor Sonde (YSI 6600, Yellow Springs Instruments, Yellow Springs, USA). Meanwhile, the collected seawater from these two layers were pre-filtered through 0.7 µm pore-sized glass fiber filters (25 mm, Whatman GF/F), and dispensed into 50 mL polycarbonate bottles and immediately stored at −20 • C for later nutrients analysis. After returning to laboratory, the ammonium (NH 4 + ), nitrate (NO 3 − ), nitrite (NO 2 − ), phosphate (PO 4 3− ) and silicate (SiO 3 2− ) concentrations were measured using an Automatic Discrete Analyzer (CleverChem Anna, DeChem-Tech. GmbH, Hamburg, Germany), and the N:P ratio was calculated as (NH 4 + + NO 3 − + NO 2 − ): PO 4 3− . The tide height was obtained from the National Marine Information Center of China (http://global-tide.nmdis.org.cn/, accessed at 20 August 2021).

Chlorophyll a (Chl a) Measurement
Every 2 h, 500 mL seawater from the surface and bottom layers were sequentially filtered through 20 µm pore-sized nylon-net filters (25 mm, Merck Millipore Ltd., Carrigtwohill, USA) and 0.7 µm pore-sized GF/F filters. The filters were then wrapped in aluminum foil, immediately frozen and stored at −20 • C for later analysis. After returning to the laboratory, these filters with cell assemblies were extracted overnight in a 5 mL magnesium carbonate saturated 90% acetone (v/v) at 4 • C in the dark. The extraction was then fluorescently measured with a Turner Designs 10 fluorometer after 10 min of centrifugation at 3500 rpm. Chl a concentration was calculated following [31], and the total Chl a was calculated by summing these two cell-sized fractions.

Fluorescence Measurement
Every 2 h, 200-300 mL of the collected seawater from surface and bottom layers were filtrated through a 20 µm pore-sized nylon-net filters by gravity. After this, the filtrate was vacuum-filtrated through 0.45 µm pore-sized polycarbonate filters (25 mm, Millipore, Burlington, MA, USA) under the pressure of~0.2 MPa. Whether this pressure was damaging the cells was not checked under microscopy, although such an effect is limited according to [32]. Phytoplankton cells reposed on the filters were resuspended in 10 mL of 0.45 µm pore-sized filtered seawater to obtain the concentrated large (micro-, >20 µm) and small (piconano-, <20 µm) cell-sized phytoplankton assemblies. Such a separation is operational and used extensively [7,27,28], although the shortcomings are obvious, e.g., some chain-forming species default to micro-cells and some other smaller cells are missing under vacuum. The concentrated samples were then dark-acclimated for 5 min in a 15 mL chamber of fast repetition rate fluorometer (FRRf), coupled to a FastAct base unit (Fast Ocean, Chelsea Technologies Group, Ltd., West Molesey, UK) where temperature was maintained at that of the field conditions, in order to oxidate the electron transport chains and relax non-photochemical quenching (NPQ) [33] before measuring chlorophyll fluorescence. After the dark-acclimation, the FRRf was activated with a single turnover protocol that consisted of 100 saturation flashets of 1 µs duration with 1 µs pitch, followed by 40 relaxation flashets with 60 µs pitch [34]. For each sample, there was a series of 10 actinic light exposures that varied from 0 to 2000 µmol photons m −2 s −1 and lasted for 60 s for each. The actinic light was delivered from a blue excitation LED (450 nm) in the FastAct instrument. From one to another light exposure step, there was a 20 s dark interval, and the first light step lasted for 120 s as phytoplankton cells need longer time to adapt to the initial transition from a dark to illuminated state [35]. The fluorescence yield from each step was recorded and averaged from 40 consecutive acquisitions. The minimum and maximum fluorescence yields in the dark-(F O , F M ) and light-regulated states (F ' , F ' M ) and the absorption cross-section of PS II photochemistry in the dark (σ PS II , nm 2 ) and light states (σ ' PS II , nm 2 ) were derived from the saturation phase of fluorescence transient with the biophysical model of [34]. The chlorophyll fluorescence was also adjusted by subtracting the fluorescence of GF/F filter-filtrated seawater to eliminate the influence of background fluorescence [36]. Photochemical PS II quantum yields (F V /F M , F q /F ' M ) in the dark-and light-state were calculated as [37]: The fluorescence light curve (FLC)-derived light utilization efficiency (α) and saturation irradiance (E K , µmol photons m −2 s −1 ) were calculated as [38,39]: where E indicates the actinic light. Then, the maximal relative electron transfer rate (rETR max ) was calculated as:

Statistical Analysis
All statistical analyses and figures were performed using R software [40]. Nonlinear regression for FLC data were conducted using the "fitWebb" function in package "phytotools" with a Nelder-Mead algorithm [41]. All of the biological and environmental parameters were centered and standardized for redundancy analysis (RDA). The RDA was performed using the "rda" function in package "vegan" to detect the effects of environmental variables on the biological or photophysiological parameters of small and large phytoplankton assemblies [42]. A paired t-test was applied to detect the significant difference, if holding an assumption of normality; otherwise, Wilcoxon signed rank exact test was applied. The significance level was set at 0.05.
In addition, the atmospheric solar photosynthetically active radiation (PAR, µmol photons m −2 s −1 ) in the surface (~0.2 m) and bottom layers (~3.5 m) of the experimental site was derived from [43] with an attenuation coefficient of 0.58 m −1 in the coastal waters of the South China Sea [44].

Results
On experimental days (25)(26), the tidal heights of the sampling site ranged from 0.32 to 1.87 m, and the PAR exceeded 1500 and 200 µmol photons m −2 s −1 at noontime in the surface and bottom layers, respectively ( Figure 2A). The seawater temperature varied from 21.2 to 22.5 • C at the surface, similar to the bottom layer; and the salinity ranged from 24.9 to 26.6 at the surface, being~10% lower than the bottom layer ( Figure 2B). The pH varied from 7.67 to 7.92 at the surface, being significantly lower than the bottom layer (i.e., 7.73-7.98) (t 17 = 15.64, p < 0.01) ( Figure 2C). In addition, the dissolved inorganic nitrogen (NH 4 parameters were centered and standardized for redundancy analysis (RDA). The RDA was performed using the "rda" function in package "vegan" to detect the effects of environmental variables on the biological or photophysiological parameters of small and large phytoplankton assemblies [42]. A paired t-test was applied to detect the significant difference, if holding an assumption of normality; otherwise, Wilcoxon signed rank exact test was applied. The significance level was set at 0.05. In addition, the atmospheric solar photosynthetically active radiation (PAR, µmol photons m −2 s −1 ) in the surface (~0.2 m) and bottom layers (~3.5 m) of the experimental site was derived from [43] with an attenuation coefficient of 0.58 m −1 in the coastal waters of the South China Sea [44].

Results
On experimental days (25)(26), the tidal heights of the sampling site ranged from 0.32 to 1.87 m, and the PAR exceeded 1500 and 200 µmol photons m −2 s −1 at noontime in the surface and bottom layers, respectively ( Figure 2A). The seawater temperature varied from 21.2 to 22.5 °C at the surface, similar to the bottom layer; and the salinity ranged from 24.9 to 26.6 at the surface, being ~10% lower than the bottom layer ( Figure  2B). The pH varied from 7.67 to 7.92 at the surface, being significantly lower than the bottom layer (i.e., 7.73-7.98) (t17 = 15.64, p < 0.01) ( Figure 2C). In addition, the dissolved inorganic nitrogen (NH4 + + NO3 − + NO2 − ), phosphate (PO4 3− ) and silicate (SiO3 2− ) concentrations that varied respectively from 7.76 to 23.62 µM, from 0.03 to 0.50 µM and from 16.25 to 52.64 µM showed no clear daily change pattern, as well as insignificant difference between the surface and bottom layers ( Figure S1). Phytoplankton biomass (Chl a) of the experimental site displayed a similar daily variation between the surface and bottom layers, i.e., decreased from morning to midday, and then increased to maximal value at dusk, and again decreased gradually to a minimal value at midnight, followed by an increase to the next morning ( Figure 3A). The surface Chl a concentration (i.e., 0.92-5.13 µg L −1 ) was significantly lower than that of the bottom layer (i.e., 1.83-6.84 µg L −1 ) (t17 = −4.70, p < 0.01), especially during daytime. Moreover, piconano-Chl a (<20 µm) accounted for 72% (mean value) of the total Chl a, with no significant difference between the surface and bottom layers ( Figure 3). According to [27], diatoms, e.g., Chaetoceros socialis, Rhizosolenia sp. and Nitzschia sp., and dinoflagellates, e.g., Scrippsiella trochoidea, generally dominated the experimental area during the spring and summer periods. Phytoplankton biomass (Chl a) of the experimental site displayed a similar daily variation between the surface and bottom layers, i.e., decreased from morning to midday, and then increased to maximal value at dusk, and again decreased gradually to a minimal value at midnight, followed by an increase to the next morning ( Figure 3A). The surface Chl a concentration (i.e., 0.92-5.13 µg L −1 ) was significantly lower than that of the bottom layer (i.e., 1.83-6.84 µg L −1 ) (t 17 = −4.70, p < 0.01), especially during daytime. Moreover, piconano-Chl a (<20 µm) accounted for 72% (mean value) of the total Chl a, with no significant difference between the surface and bottom layers ( Figure 3). According to [27], diatoms, e.g., Chaetoceros socialis, Rhizosolenia sp. and Nitzschia sp., and dinoflagellates, e.g., Scrippsiella trochoidea, generally dominated the experimental area during the spring and summer periods. Microorganisms 2022, 9, x FOR PEER REVIEW 6 of 13 Daily variations of the maximum PS II photochemical quantum yield (FV/FM) and the functional absorption cross-section of PS II photochemistry (σPS II) of both piconano-and micro-phytoplankton assemblies are shown in Figure 4. The FV/FM of the piconano-cell assembly at the surface decreased from 0.43 ± 0.02 to 0.35 ± 0.03 from morning to noontime, then increased to 0.41 ± 0.01 at dusk, and again decreased to 0.37 ± 0.01 at midnight, followed by an increase to next morning ( Figure 4A). The FV/FM of the micro-cell assembly displayed a greater reduction at noontime than the piconano-ones, indicating the more light-caused photoinhibition. In the bottom layer, the FV/FM of both the piconano-and micro-cell assemblies were significantly higher than that of the surface (piconano-, t17 = 5.65, p < 0.01; micro-, t17 = 6.07, p < 0.01), in particular at daytime, and showed less daily variation. The σPS II, a measure of quantum yield for PS II photochemistry, showed a similar pattern as the FV/FM during daytime; during the nighttime, however, the σPS II of the microcell assemblies from both the surface and bottom layers decreased from dusk until the next morning (surface, R 2 = 0.76, p < 0.05; bottom, R 2 = 0.63, p < 0.05), which did not occur in the piconano-cell assemblies ( Figure 4C). Moreover, the σPS II of the piconano-cell assembly that varied from 3.15 to 4.13 nm 2 was significantly higher than that of the microcell assemblies (i.e., 2.71-3.88 nm 2 ), no matter in what time and depth (surface, t17 = 6.67, p < 0.01; bottom, t17 = 5.88, p < 0.01) ( Figure 4D). Consistently, the FLC-derived light utilization efficiency (α) of both types of cell assemblies from both layers showed the same daily change pattern in terms of the FV/FM (Figure 5A), while the saturation irradiance (EK) and maximal rETR (rETRmax) mirrored the field light fluctuations ( Figure 5C,E). The α of the piconano-cell assembly from both the surface and bottom layers was lower than the micro-cell assemblies under low light or dark conditions (t25 = −2.62, p < 0.05), but not under high light ( Figure 5B); while the EK and rETRmax were always lower (t17 = 4.37-8.57, p < 0.01) ( Figure 5D,F). Daily variations of the maximum PS II photochemical quantum yield (F V /F M ) and the functional absorption cross-section of PS II photochemistry (σ PS II ) of both piconanoand micro-phytoplankton assemblies are shown in Figure 4. The F V /F M of the piconanocell assembly at the surface decreased from 0.43 ± 0.02 to 0.35 ± 0.03 from morning to noontime, then increased to 0.41 ± 0.01 at dusk, and again decreased to 0.37 ± 0.01 at midnight, followed by an increase to next morning ( Figure 4A). The F V /F M of the micro-cell assembly displayed a greater reduction at noontime than the piconano-ones, indicating the more light-caused photoinhibition. In the bottom layer, the F V /F M of both the piconanoand micro-cell assemblies were significantly higher than that of the surface (piconano-, t 17 = 5.65, p < 0.01; micro-, t 17 = 6.07, p < 0.01), in particular at daytime, and showed less daily variation. The σ PS II , a measure of quantum yield for PS II photochemistry, showed a similar pattern as the F V /F M during daytime; during the nighttime, however, the σ PS II of the micro-cell assemblies from both the surface and bottom layers decreased from dusk until the next morning (surface, R 2 = 0.76, p < 0.05; bottom, R 2 = 0.63, p < 0.05), which did not occur in the piconano-cell assemblies ( Figure 4C). Moreover, the σ PS II of the piconano-cell assembly that varied from 3.15 to 4.13 nm 2 was significantly higher than that of the micro-cell assemblies (i.e., 2.71-3.88 nm 2 ), no matter in what time and depth (surface, t 17 = 6.67, p < 0.01; bottom, t 17 = 5.88, p < 0.01) ( Figure 4D). Consistently, the FLC-derived light utilization efficiency (α) of both types of cell assemblies from both layers showed the same daily change pattern in terms of the F V /F M (Figure 5A), while the saturation irradiance (E K ) and maximal rETR (rETR max ) mirrored the field light fluctuations ( Figure 5C,E). The α of the piconano-cell assembly from both the surface and bottom layers was lower than the micro-cell assemblies under low light or dark conditions (t 25 = −2.62, p < 0.05), but not under high light ( Figure 5B); while the E K and rETR max were always lower (t 17 = 4.37-8.57, p < 0.01) ( Figure 5D    When the pooled F V /F M from the surface and bottom layers was plotted against the Chl a biomass, a positive correlation occurred in both the piconano-(R 2 = 0.32, p < 0.01) and micro-cell assemblies (R 2 = 0.21, p < 0.01) ( Figure 6). However, the increased degree, as indicated by the slope, was~4-fold higher in micro-than piconano-cell assemblies (i.e., 0.063 ± 0.020 vs. 0.016 ± 0.004) (p < 0.01), indicating that the large phytoplankton cells are more susceptible to the change in Chl a biomass that often mirrors the environmental changes.
Based on the RDA results, phytoplankton biomass (Chl a) and photosynthetic parameters (F V /F M , α and rETR max ) of both piconano-and micro-cell assemblies were positively correlated to the temperature, salinity and pH, as well as the N:P ratio (Figure 7). Moreover, the F V /F M , α and σ PS II were negatively correlated to the solar PAR, indicating the photoinactivation of PS II under high light. and absorption cross section for PSII photochemistry (C, σPSII, nm 2 ) of piconano-(<20 µm) and micro-cell assemblies (>20 µm) in surface and bottom layers of experimental site, as well as the comparisons of (B) FV/FM or (D) σPS II between piconano-and micro-cell assemblies. Grey shadows in panels A and C indicate the nighttime. Solid lines in panels B and D show the linear regression with shadow indicating 95% confidence interval; and the color in symbol indicates the ln(PAR + 1) wherein the PAR is the cell-received irradiance in field condition.  When the pooled FV/FM from the surface and bottom layers was plotted against the Chl a biomass, a positive correlation occurred in both the piconano-(R 2 = 0.32, p < 0.01) and micro-cell assemblies (R 2 = 0.21, p < 0.01) ( Figure 6). However, the increased degree, as indicated by the slope, was ~4-fold higher in micro-than piconano-cell assemblies (i.e., 0.063 ± 0.020 vs. 0.016 ± 0.004) (p < 0.01), indicating that the large phytoplankton cells are more susceptible to the change in Chl a biomass that often mirrors the environmental changes. Based on the RDA results, phytoplankton biomass (Chl a) and photosynthetic parameters (FV/FM, α and rETRmax) of both piconano-and micro-cell assemblies were positively correlated to the temperature, salinity and pH, as well as the N:P ratio (Figure 7). Moreover, the FV/FM, α and σPS II were negatively correlated to the solar PAR, indicating the photoinactivation of PS II under high light.  . Redundancy analysis (RDA) for correlation of pooled photosynthetic parameters and environmental variables for piconano-(<20 µm) and micro-cell assemblies (>20 µm). Constrained axes RDA1 and RDA2 account for 15.36% and 11.16% of total variance, respectively. The color in the symbol indicates the ln(PAR + 1) wherein the PAR is the cell-received irradiance in field condition. Temp, temperature; Sal, salinity.

Discussion
Both laboratory and field studies have evidenced that phytoplankton cells exhibit the diel rhythm in photosynthetic behaviors [14,17,19,21]. In this study, we found piconanophytoplankton cells displayed higher photosynthetic capacity and less photoinhibition under local noontime light, as compared to their micro-cell counterparts. Such a differential degree of the daily variation in photosynthetic parameters between these two differently cell sized phytoplankton assemblies indicated that the cell size range may alter the diel rhythm of their physiological behaviors under field conditions. Moreover, the photosynthetic efficiency of micro-cells varied more quickly to the Chl a change than the piconano-cells, suggesting that the larger phytoplankton cells may be more suitable to grow in the variable environments of, e.g., the Daya Bay, through mediating the activity of photosystem in a timely fashion.
Chl a biomass in both surface and bottom layers displayed great variation throughout the days (Figure 3), and positively correlated to the tidal heights (surface, R 2 = 0.31, p < 0.05; bottom, R 2 = 0.41, p < 0.01). The tidal cycle, i.e., the ebb and flow, is generally believed to alter the movement of the water body, which in turn can change its physio-chemical and biological properties by mixing. Such mixing can also inevitably alter the abundance and species composition of phytoplankton through mixing the water bodies that originally contain different cell biomass and species, and/or through altering the physiochemical properties and thus the cell growth [45,46]. The tide rise-induced seawater intrusion often brings more clear water into the sampling site, and as a result the phytoplankton within this water would be higher if considering the presence of more underwater solar energy and plentiful nutrients ( Figure S1) [23] that would endow them to grow faster. Therefore, the seawater intrusion may have contributed to the higher Chl a biomass ( Figure 3A). What is more, the clearer water body in the sampling site during high tide may also benefit the growth of phytoplankton cells, but this effect on the Chl a biomass Figure 7. Redundancy analysis (RDA) for correlation of pooled photosynthetic parameters and environmental variables for piconano-(<20 µm) and micro-cell assemblies (>20 µm). Constrained axes RDA1 and RDA2 account for 15.36% and 11.16% of total variance, respectively. The color in the symbol indicates the ln(PAR + 1) wherein the PAR is the cell-received irradiance in field condition. Temp, temperature; Sal, salinity.

Discussion
Both laboratory and field studies have evidenced that phytoplankton cells exhibit the diel rhythm in photosynthetic behaviors [14,17,19,21]. In this study, we found piconanophytoplankton cells displayed higher photosynthetic capacity and less photoinhibition under local noontime light, as compared to their micro-cell counterparts. Such a differential degree of the daily variation in photosynthetic parameters between these two differently cell sized phytoplankton assemblies indicated that the cell size range may alter the diel rhythm of their physiological behaviors under field conditions. Moreover, the photosynthetic efficiency of micro-cells varied more quickly to the Chl a change than the piconano-cells, suggesting that the larger phytoplankton cells may be more suitable to grow in the variable environments of, e.g., the Daya Bay, through mediating the activity of photosystem in a timely fashion.
Chl a biomass in both surface and bottom layers displayed great variation throughout the days (Figure 3), and positively correlated to the tidal heights (surface, R 2 = 0.31, p < 0.05; bottom, R 2 = 0.41, p < 0.01). The tidal cycle, i.e., the ebb and flow, is generally believed to alter the movement of the water body, which in turn can change its physio-chemical and biological properties by mixing. Such mixing can also inevitably alter the abundance and species composition of phytoplankton through mixing the water bodies that originally contain different cell biomass and species, and/or through altering the physio-chemical properties and thus the cell growth [45,46]. The tide rise-induced seawater intrusion often brings more clear water into the sampling site, and as a result the phytoplankton within this water would be higher if considering the presence of more underwater solar energy and plentiful nutrients ( Figure S1) [23] that would endow them to grow faster. Therefore, the seawater intrusion may have contributed to the higher Chl a biomass ( Figure 3A). What is more, the clearer water body in the sampling site during high tide may also benefit the growth of phytoplankton cells, but this effect on the Chl a biomass would be eliminated in such an hourly scale. The nutrients concentration also plays a pivotal role in modulating phytoplankton growth and even blooms [47], but such a nutrient-induced activation for cell growth could not explain the Chl a being so altered in an hourly scale either [48]. Finally, the small cell-sized phytoplankton prevailed in the experimental area, consistent with previous studies [17,27,28]. It seemed to be contradicted that larger cells can regulate the PS II activity faster than smaller ones ( Figure 6) and are thus more suitable to grow in the varied environments of coastal regions. This may be explained that the sampling site is located in shellfish culture areas [49], and the selective feedings of shellfish may have altered the community structure of phytoplankton assemblages [49,50].
The photosynthetic parameter of F V /F M usually decreases under stressful high light, caused by the photoinactivation of PS II [13,51,52]; therefore, it is often used as an indicator of photosynthetic capacity of phytoplankton cells [8,17]. Consistently, the F V /F M was lower at noontime than the other times ( Figure 4A); while this decline was less in piconano-than in micro-cell assemblies, which, together with a similar decline of the FLC-derived light utilization efficiency (α, Figure 5A), implies that smaller cells are more resistant to high light. However, this finding is contrary to the previously-established concept that the lower package effect of smaller cells can render them more susceptible to high light [8,53]. Such a contradiction may be explained by the fact that the smaller cells have higher damage repair capacity, and, considering their higher metabolic rate, they may have actively energized the repair process under stressful high light [52], thus maintaining the higher photosynthetic efficiency than their larger counterparts ( Figures 4A and 5A). Moreover, smaller cells often have higher antioxidant activity that may help them to relieve the oxidative damage triggered by reactive oxygen species (ROS) [16], thus enhancing their high light resistance. More interestingly, our results also showed that the F V /F M of the piconano-cell assembly varied more slowly to the Chl a biomass change than the micro-cell assembly (Figure 6), which indicates the slower reactiveness of smaller cells to environmental changes, and by contrast, the larger cells can adjust to the activity of PS II in a timely fashion and may be thus more suitable to grow in the variable environments of the Daya Bay. This was also supported by the higher changing amplitude of saturation irradiance (E K ) and electron transport rate (rETR max ) in micro-cell assembly than in piconano-cell assemblies ( Figure 5C,E). Furthermore, the higher σ PS II of piconano-cell assembly than micro-cell assemblies ( Figure 4C,D) indicates the higher light-harvesting ability of smaller cells, also supporting their higher light utilization efficiency ( Figure 5A) and photosynthetic rate ( Figure 5E). It may benefit them in turbid water too, thus maintaining their high abundance in the Daya Bay, as found in this study ( Figure 2) or in others [17,27,28].

Conclusions
In this study, we found that the tide's coming in and going out co-varied with the biomass and community structure of phytoplankton assemblages from the experimental site through moving the water bodies. We also found the piconano-phytoplankton assembly exhibited higher photosynthetic efficiency, as well as more tolerance to high light compared to their micro-cell counterparts. Moreover, we found the piconano-cell assembly exhibited less daily variation in the photosynthetic parameters and slower physiological reactiveness to environmental changes, as compared to micro-cell assemblies. In particular, our results from in-situ conditions, together with those from laboratory conditions [13,14,22], indicated that the varied community structure modulates the diel rhythm of the photophysiological behaviors of phytoplankton assemblages in the Daya Bay. In addition, our findings in the eutrophic regions may have been tempered by considering that the concentration of the 0.45 µm pore-size filters under vacuum for piconano-cell assembly may not cover all smaller cells of, e.g., Synechoccocus, and such a tempered effect would be even more evident in the piconano-cell dominated oligotrophic oceans [7] or summer season of the Daya Bay [27]. Therefore, further studies should be performed in other nutrient-status regions or other seasons to generalize an implication being suitable for broader frames.